nts 
  
  
  
  
  
  
  
prt ed 2 2 2 
s Bs 2 (4° sy) x EE à b(pr2d?) | 4 42 + 
‚em 756° E b) a DUnEMd)| b. 
- xy" «X. + (2x-b)ba“ | © _3 + (2x-b)y“ x^ 35 1 b(b^424^) i 
b 2 20:34 d wq ln 2 20544.) 
2 
20 au 
80° 20(b^ -4a^) 
Hern X 142 
20 16d ; a 44 
For wide angle photographs with normal overlap (60%) and Sg = S we find 
the following mean values for the neat model area 
M = 1.15 Rs 
x c 0 
M:lAD. 
y c 0 
: h 
M,- 1.8— 59 
3.2 Large elevation differences. 
  
For models of terrain with large elevation differences general formulas for 
the accuracy can hardly be derived. But for each individual case the expres- 
sions for the accuracy to be expected can be demonstrated in formulas, 
which are derived in a similar manner as those which are demonstrated 
above for flat ground. 
For the test area Monti di Revóira such formulas have been derived for the 
actual models. Due to the rather regular inclination of the ground and the 
positions and number of control points the expressions for the individual 
models became similar. 
In the diagrams 1.1 - 1.3 below the distributions of the standard error of 
the final coordinates and elevations of the model 2412-2413 are demonstrated 
for ss, = s, - 1. 
0 0 
The mean values of the standard errors for the neat models are 
h 
M = 1.5 = Sp 
8 h 
M 1 5-780 
: h 
M = 2.0 78g 
his the average flying altitude and c is the camera constant. 
Consequently 2 is the average image scale factor 5.